6994
Langmuir 2003, 19, 6994-7002
Photolithographic Polymerization of
Diacetylene-Containing Phospholipid Bilayers Studied by
Multimode Atomic Force Microscopy
Kenichi Morigaki,*,†,§,| Holger Schönherr,*,†,‡,⊥ Curtis W. Frank,†,‡ and
Wolfgang Knoll†,‡,§
NSF MRSEC Center on Polymer Interfaces and Macromolecular Assemblies (CPIMA),
Department of Chemical Engineering, Stanford University, Stanford, California 94305-5025,
and Max-Planck-Institute for Polymer Research, 55128 Mainz, Germany
Received January 16, 2003. In Final Form: January 28, 2003
Photopolymerization of the diacetylene-containing phospholipid 1,2-bis(10,12-tricosadiynoyl)-sn-glycero3-phosphocholine (1) in substrate-supported planar lipid bilayers (SPBs) has been studied by using multimode
atomic force microscopy (AFM). Monolayers and bilayers of 1 have been transferred onto glass substrates
by the Langmuir-Blodgett (LB) and Langmuir-Schaefer (LS) techniques, respectively. Bilayers of
monomeric 1 were removed easily from the substrate surface by detergent solutions (0.08 M sodium dodecyl
sulfate (SDS)). Once they were photopolymerized by UV light, the bilayers became resistant toward detergent
solubilization or exposure to air. High-resolution AFM images revealed the molecular packing of the
hydrocarbon tails and the headgroups for monolayers and bilayers, respectively. The lattice spacing of
photopolymerized 1 bilayers was similar to that of monomeric bilayers, suggesting that the polymerization
proceeds without disruption of the bilayer structure. Patterned SPBs of poly-1 were obtained by UV
photolithography and selective removal of the protected monomeric bilayers. AFM revealed that the
photopolymerization proceeds heterogeneously within the bilayers, depending on the UV irradiation dose.
At longer UV exposure times, side reactions and/or reorganization of the molecules within the bilayer may
occur.
Introduction
The modification of solid surfaces with biological
molecules is currently studied with the aim of developing
biomimetic interfaces that will play key roles in numerous
biomedical and environmental applications. Among various proposed formats of functionalized interfaces, substrate-supported planar lipid bilayers (abbreviated as
SPBs in the following) provide a unique possibility for
reconstituting cellular membranes on a solid surface.1,2
SPB micropatterning has attracted considerable attention
because it allows the creation of designed microarrays of
biological materials and should facilitate various new
applications, such as high throughput drug screening.3-7
We have recently reported a novel SPB micropatterning
strategy based on the lithographic photopolymerization
* Corresponding authors.
† CPIMA.
‡ Stanford University.
§ Max-Planck-Institute for Polymer Research.
| Present address: National Institute of Advanced Industrial
Science and Technology (AIST), Ikeda 563-8577, Japan. Fax: +8172-751-9628. E-mail: [email protected].
⊥ Present address: University of Twente, Faculty of Science and
Technology and MESA+ Research Institute, Department of Material
Science and Technology of Polymers, P.O. Box 217, 7500 AE
Enschede, The Netherlands. Fax: +31-53-489-3823. E-mail:
[email protected].
(1) Sackmann, E. Science (Washington, D.C.) 1996, 271, 43-48.
(2) Plant, A. Langmuir 1993, 9, 2764-2767.
(3) Groves, J. T.; Ulman, N.; Boxer, S. G. Science (Washington, D.C.)
1997, 275, 651-653.
(4) Jenkins, A. T. A.; Boden, N.; Bushby, R. J.; Evans, S. D.; Knowles,
P. F.; Miles, R. E.; Ogier, S. D.; Schönherr, H.; Vancso, G. J. J. Am.
Chem. Soc. 1999, 121, 5274-5280.
(5) Cremer, P. S.; Yang, T. J. Am. Chem. Soc. 1999, 121, 8130-8131.
(6) Groves, J. T.; Mahal, L. K.; Bertozzi, C. R. Langmuir 2001, 17,
5129-5133.
(7) Fang, Y.; Frutos, A. G.; Lahiri, J. J. Am. Chem. Soc. 2002, 124,
2394-2395.
of the diacetylene phospholipid 1,2-bis(10,12-tricosadiynoyl)-sn-glycero-3-phosphocholine (1).8,9 The fabrication process comprises four steps (a schematic illustration
is given in Figure 1): (i) formation of a monomeric bilayer
on a solid substrate, (ii) lithographic photopolymerization
by UV light, (iii) removal of the unpolymerized bilayers,
and (iv) refilling the empty areas with new lipid bilayers.
The lipid bilayers, which are incorporated at the last step,
retain some characteristic features of native cellular
membranes (e.g., lateral fluidity) and are intended to be
used for further biological applications. Since the defined
matrix of the polymerized lipid bilayer templates is
prepared in the first three steps, these steps are essential
for the current micropatterning strategy.
We have employed diacetylene phospholipid 1 for the
photopolymerization of lipid bilayers (Figure 2). Photopolymerization of this and similar diacetylenes proceeds
in the solid state (topochemical polymerization) and results
in rigid films.10-12 Preorganization of the diacetylene
moieties can be achieved in Langmuir-Blodgett (LB) films
and self-assembled monolayers (SAMs). These assemblies
based on diacetylene-containing amphiphiles have been
investigated as potential photoresist materials.13-16
(8) Morigaki, K.; Baumgart, T.; Offenhäusser, A.; Knoll, W. Angew.
Chem., Int. Ed. Engl. 2001, 40, 172-174.
(9) Morigaki, K.; Baumgart, T.; Jonas, U.; Offenhäusser, A.; Knoll,
W. Langmuir 2002, 18, 4082-4089.
(10) Wegner, G. Makromol. Chem. 1972, 154, 35-48.
(11) Polydiacetylenes; Cantow, H.-J., Ed.; Springer-Verlag: Berlin,
1984.
(12) Polydiacetylenes: Synthesis, Structure and Electronic Properties;
Bloor, D., Chance, R. R., Eds.; Martinus Nijhoff Publishers: Dordrecht/
Boston/Lancaster, 1985.
(13) Barraud, A. Thin Solid Films 1983, 99, 317-321.
(14) Batchelder, D. N.; Evans, S. D.; Freeman, T. L.; Häussling, L.;
Ringsdorf, H.; Wolf, H. J. Am. Chem. Soc. 1994, 116, 1050-1053.
(15) Chan, K. C.; Kim, T.; Schoer, J. K.; Crooks, R. M. J. Am. Chem.
Soc. 1995, 117, 5875-5876.
10.1021/la034078f CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/19/2003
Photopolymerization of Diacetylene Phospholipid
Langmuir, Vol. 19, No. 17, 2003 6995
thermore, diacetylene-containing phospholipids have been
polymerized in lipid vesicles (liposomes) with the aim of
producing mechanically robust drug carrier capsules.22-26
It has been also reported that some chiral diacetylene
phospholipids form unique tubular structures in water if
they are incubated below the solid-fluid phase transition
temperature (Tc) of the bilayers.27,28 Various applications
of these tubules have also been proposed.29
In previous work, we have obtained information on the
successful photopatterning strategy mainly relying on the
specific fluorescence of the polymer backbones.8,9 Herein
we report on the complementary real-space characterization of the morphology and nanometer-scale order of the
layers by multimode atomic force microscopy (AFM) during
all steps of the fabrication process. Although polymerized
monolayers of diacetylene-containing adsorbates have
been studied extensively by scanning probe methods,30-33
this is the first report of an AFM investigation of the
nanometer-scale structure and morphology of unpolymerized and polymerized bilayers of diacetylene phospholipids. Our observations shed light on the underlying
processes of our micropatterning strategy and may lead
to the development of various technical improvements.
Materials and Methods
Figure 1. Schematic of the strategy employed for the micropatterning of lipid bilayer membranes.
Figure 2. Polymerizable diacetylene phospholipid 1 and the
polymerization scheme.
Poly(diacetylenes) form long conjugated sequences of
ene-yne backbones, which absorb UV/visible light
strongly. This property has attracted much attention
because of potential optical and electronic applications.
In particular, the reversible and nonreversible color
changes due to the structural transition between blue and
red polymorphs of poly(diacetylenes) have been studied
extensively for colorimetric sensor applications.17-21 Fur(16) Mowery, M. D.; Smith, A. C.; Evans, C. E. Langmuir 2000, 16,
5998-6003.
(17) Charych, D. H.; Nagy, J. O.; Spevak, W.; Bednarski, M. D. Science
(Washington, D.C.) 1993, 261, 585-588.
(18) Okada, S.; Peng, S.; Spevak, W.; Charych, D. Acc. Chem. Res.
1998, 31, 229-239.
(19) Jonas, U.; Shah, K.; Norvez, S.; Charych, D. H. J. Am. Chem.
Soc. 1999, 121, 4580-4588.
(20) McQuade, D. T.; Pullen, A. E.; Swager, T. M. Chem. Rev. 2000,
100, 2537-2574.
(21) Kolusheva, S.; Boyer, L.; Jelinek, R. Nature Biotechnol. 2000,
18, 225-227.
Materials. Diacetylene phospholipid (1), 1,2-bis(10,12-tricosadiynoyl)-sn-glycero-3- phosphocholine, was purchased from
Avanti Polar Lipids (Alabaster, AL); sodium dodecyl sulfate (SDS)
was purchased from Fluka (Buchs, Switzerland); all other
chemicals were purchased from Sigma (St. Louis, MO). These
commercially obtained chemicals were used without further
purification. As substrates for the SPBs, we have used either
microscopy glass slides (Menzel, Braunschweig, Germany),
polished quartz (Hellma, Müllheim, Germany), or silicon wafers.
Substrate Cleaning. The substrates were cleaned first with
a commercial detergent solution (0.5% Hellmanex/water, Hellma,
Mühlheim, Germany), rinsed with deionized water, treated in
a warm solution of 28% NH4OH/30% H2O2/H2O (1:1:5) (65 °C for
15 min), rinsed extensively with deionized water, and then dried
in a vacuum oven at 110 °C. This protocol resulted in clean and
hydrophilic surfaces for the adsorption of lipid bilayer membranes.
Preparation of Supported Planar Bilayers. Monolayers
and bilayers of monomeric 1 were deposited onto solid substrates
from the air/water interface using a KSV5000 Langmuir trough
(KSV Instruments, Helsinki, Finland). Diacetylene lipid 1 was
spread from a chloroform solution. The lipids formed stable
monolayers at the air/water interface up to a surface pressure
of 40 mN/m at 20 °C. The monolayers were transferred onto solid
substrates at 35 mN/m (fully condensed state). The first
monolayer was deposited by dipping and withdrawing the
substrate vertically (Langmuir-Blodgett (LB) method). The
second leaflet was deposited onto the hydrophobic surface of the
(22) Hub, H.-H.; Hupfer, B.; Koch, H.; Ringsdorf, H. Angew. Chem.,
Int. Ed. Engl. 1980, 19, 938-940.
(23) Johnston, D. S.; McLean, L. R.; Whittam, M. A.; Clark, A. D.;
Chapman, D. Biochemistry 1983, 22, 3194-3202.
(24) Ringsdorf, H.; Schlarb, B.; Venzmer, J. Angew. Chem., Int. Ed.
Engl. 1988, 27, 113-158.
(25) Freeman, F. J.; Chapman, D. In Liposomes as Drug Carriers;
Gregoriadis, G., Ed.; John Wiley & Sons: New York, 1988; pp 821-839.
(26) Mueller, A.; O’Brien, D. F. Chem. Rev. 2002, 102, 727-757.
(27) Yager, P.; Schoen, P. E. Mol. Cryst. Liq. Cryst. 1984, 106, 371381.
(28) Schnur, J. M.; Ratna, B. R.; Selinger, J. V.; Singh, A.; Jyothi,
G.; Easwaran, K. R. K. Science (Washington, D.C.) 1994, 264, 945-947.
(29) Schnur, J. M. Science (Washington, D.C.) 1993, 262, 1669-1676.
(30) Marti, O.; Ribi, H. O.; Drake, B.; Albrecht, T. R.; Quate, C. F.;
Hansma, P. K. Science (Washington, D.C.) 1988, 239, 50-52.
(31) Nelles, G.; Schönherr, H.; Jaschke, M.; Wolf, H.; Schaub, M.;
Küther, J.; Tremel, W.; Bamberg, E.; Ringsdorf, H.; Butt, H. J. Langmuir
1998, 14, 808-815.
(32) Sasaki, D. Y.; Carpick, R. W.; Burns, A. R. J. Colloid Interface
Sci. 2000, 229, 490-496.
(33) Okawa, Y.; Aono, M. Nature (London) 2001, 409, 683-684.
6996
Langmuir, Vol. 19, No. 17, 2003
first monolayer by pressing the substrate horizontally through
the monolayer at the air/water interface and dropping it into the
subphase (Langmuir-Schaefer (LS) method). The deposition of
the second monolayer was typically made 1-2 h after the first.
After the deposition of the second monolayer, the samples were
collected from the trough and stored in deionized water (in the
dark).
Photopolymerization of Bilayers. The polymerization was
conducted in a closed system that comprised a water reservoir,
a pump, and a cell (ca. 4 mL volume) equipped with a quartz
window using a small low pressure mercury lamp (2 W, strong
emission band at 254 nm, UVP, Pen-Ray, Upland, CA) as the
light source. Prior to polymerization, argon-purged oxygen-free
water was circulated continuously by the pump (3.8 mL/min)
through the polymerization cell for typically 15 min.34 UV
photopatterning was achieved by illuminating the SPB through
a contact mask (gold pattern on quartz), which was mounted
onto the sample inside the cell. The distance between the UV
light source and the SPB was 5 cm.
Solubilization of Layers of Lipid 1 by Detergent. Nonpolymerized 1 molecules were removed from the substrate surface
with SDS solutions. The film thickness of monolayers and bilayers
of 1 was measured by ellipsometry in air (L116C, Gaertner,
Chicago, IL, wavelength 632.8 nm, 70° angle of incidence,
polarizer set to 45°) after treatment in various concentrations of
SDS. Silicon wafers were used as substrates for these experiments. Since the ellipsometry measurements were conducted in
air, the measured thickness of the bilayers was an average value
of the collapsed films. By contrast, the thicknesses of monolayers
were measured for intact films.
Atomic Force Microscopy. The AFM experiments were
carried out on a NanoScope III (Digital Instruments (DI), Santa
Barbara, CA) in contact mode (CM) and tapping mode (TM). For
CM-AFM V-shaped Si3N4 cantilevers with various spring constants were used (regular tips, Nanoprobes (DI), knominal ) 0.060.58 N/m; oxide sharpened tips, Olympus, knominal ) 0.02 N/m).
The triangular Si3N4 cantilevers with knominal ) 0.58 N/m were
also used for TM-AFM in water. The TM-AFM data in air were
acquired with single beam Si cantilevers (Nanosensors, Wetzlar,
Germany, ν0 ≈ 300 kHz). Measurements were performed as
specified in air (30%-40% humidity, 24 °C ambient temperature)35 or liquid (Milli-Q water or solutions of SDS in Milli-Q
water, temperature inside the liquid cell ca. 30-32 °C) using a
DI liquid cell. The scanner was protected against liquid spillage
by covering the top part with a very thin film of sealant (Parafilm,
American National Can, Neenah, WI). For most experiments
under liquid, the rubber ring was not used since the samples
were glued to the sample holder using pressure-sensitive adhesive
and since the maximum scan range without distortions due to
rubber ring flexing is not compromised. The use of pressuresensitive adhesive was necessary to mount samples onto the
AFM while keeping the surface, which was later imaged, covered
by water at all times. For this purpose, the rubber ring was
pressed under water onto the sample supported on a glass slide.
The slide and the rubber ring were taken out of solution and
placed onto a sheet of adsorbent paper to dry the bottom side of
the slide. If necessary, water was refilled into the rubber ring
using a syringe. Next, the sample was transferred onto the sample
holder, which was covered with double-sided, pressure-sensitive
adhesive. The holder was mounted onto the protected scanner
with minimal delay. Then the rubber ring was removed and
additional water was deposited onto the sample, ensuring that
the water spanned the distance between the sample and the
liquid cell of the AFM after the AFM head was put onto the
scanner. After a brief equilibration period of typically several
minutes necessary to obtain a stable photodiode reading, the
experiments were started. For high-resolution CM-AFM measurements under water, the samples were glued with epoxy
(Devcon 5 Minute Epoxy, ITW Devcon, Danvers, MA) to the
sample holder disk. Here the AFM liquid cell was used with the
(34) Oxygen had to be removed from the aqueous solution prior to
photopolymerization, since the presence of oxygen inhibited the
polymerization, presumably by quenching diacetylene radicals. Day,
D.; Ringsdorf, H. J. Polym. Sci., Polym. Lett. Ed. 1978, 16, 205-210.
(35) Due to heat generated by the AFM, this corresponds to a true
sample temperature of ca. 34 °C.
Morigaki et al.
Figure 3. TM-AFM height image of a monolayer of monomeric
1 on glass acquired in air.
rubber ring. Equilibration times of up to 6 h ensured that thermal
and instrumental drift were minimized. CM-AFM experiments
in liquid and in air were carried out with controlled forces adjusted
according to the additionally recorded force-distance plots.
Usually the lowest possible imaging force was used (<200 pN).
The operating conditions in tapping mode in liquid and in air
were also adjusted to minimal peak forces by keeping the
undamped amplitude as low as possible and maintaining a high
set-point amplitude ratio. All the images shown here were
subjected to a first-order plane-fitting procedure to compensate
for sample tilt and, if necessary, to a zero-order flattening.
Results
The characterization of the monomeric monolayers and
bilayers is described first. Following the discussion on the
solubilization procedures, the photopolymerization of
bilayers and the characterization of the obtained polymer
layers will be described.
Morphology and Molecular Order in Monomeric
Monolayers/Bilayers on Glass Substrates. Figure 3
shows a TM-AFM image of a monolayer acquired in air.
Although scattered dust particles can be recognized, there
are essentially no discernible defects present in the layer.
Monolayers of monomeric 1 were very flat; the measured
root-mean-square (rms) roughness was 0.3 ( 0.06 nm (scan
size 5 µm × 5 µm), which is comparable to the glass
substrates used in this study (0.4 ( 0.04 nm). At defect
sites or intentionally induced scratches of the monomeric
monolayer film, we measured an apparent layer thickness
of ca. 1.3 nm. This step height does not correspond directly
to the monolayer height since it is unlikely that the bare
substrate is exposed at the defects observed (vide infra).
High-resolution images recorded in contact mode in air
show a near-hexagonal periodic lattice structure with d
) 5.8 Å (corresponding to an area per molecule of ca. 29
Å2) (Figure 4a). These values compare favorably to the
spacing/area requirement of individual hydrocarbon tails.
Prolonged storage of the monomer films in air led to
reorganization of the film structure to a needlelike
morphology. These needles showed a markedly different
lattice structure at the surface as revealed by CM-AFM
in air (Figure 4b). The restructured crystals form a lattice,
which can be described as rectangular (a1 ) 4.9 Å, b1 )
7.3 Å). The area requirement of 36 Å2 per molecule suggests
a significantly different orientation of the molecules
compared to the as-transferred Langmuir monolayer
(hexagonal lattice with an area per molecule of ca. 29 Å2).
Photopolymerization of Diacetylene Phospholipid
Langmuir, Vol. 19, No. 17, 2003 6997
Figure 4. High-resolution CM-AFM images of monolayers of monomeric 1 on glass acquired in air: (a) height image of freshly
prepared sample; (b) deflection image of aged, restructured lipid layer. Insets: 2-D FFT, autocovariance filtered sections.
Figure 5. (a) CM-AFM height image of a monomeric bilayer of 1 recorded in water; (b) height profile along the white line of (a).
The step heights indicated by the markers are 5.4 and 2.8 nm, respectively.
This increase in area per molecule is consistent with an
increase in tilt angle of the molecules with respect to the
surface-normal direction.
After transfer of the second layer of 1 using the LS
technique, the AFM measurements were carried out in
water. Both TM-AFM (no data shown) and CM-AFM
revealed a smooth bilayer with scattered defects (Figure
5a). The number of defects is significantly increased
compared to the monolayer discussed above. This observation may be related to the difference in transfer
procedure (LB versus LS). The rms roughness of the
monomeric bilayer was measured to be 0.5 ( 0.23 nm.
The section analysis (Figure 5b) shows step heights of
∼3.0 and ∼5.6 nm at defect sites in the bilayer, which
agree well with the expected heights of the monolayer
and bilayer.
High-resolution CM-AFM images recorded in water
unveiled a periodic lattice structure on the nanometer
scale (Figure 6). This lattice is clearly different from the
structure observed for the monolayer (Figure 4a). From
the two-dimensional fast Fourier transform (FFT), we
calculate a rectangular lattice (a2 ) 11.3 Å, b2 ) 5.8 Å
corresponding to an area requirement of 66 Å2/molecule),
which is consistent with the packing of the lipid headgroups.
The bilayers stayed under water at all times during the
experiments and were found to be stable for a period longer
Figure 6. High-resolution CM-AFM friction image of a
monomeric bilayer of 1 acquired in water (insets: 2-D FFT,
autocovariance filtered section).
than 100 days in water. Exposing these layers to air
without rinsing led to a characteristic collapse of the
bilayer structure. A typical TM-AFM image recorded in
air is shown in Figure 7a. The step height, as measured
in cross-sectional plots (Figure 7b), was 5.5 nm. Careful
6998
Langmuir, Vol. 19, No. 17, 2003
Morigaki et al.
Figure 7. (a) TM-AFM image of collapsed monomeric bilayer of 1 imaged in air; (b) height profile along the white line in (a). The
step height indicated by the markers is 5.2 nm.
Figure 8. Dissolution of monolayers and bilayers of 1 from
oxidized silicon surface by SDS solutions. The films were
immersed in SDS solutions of various SDS concentrations for
30 min, and the remaining film was measured by null
ellipsometry.
rinsing with water removed the collapsed bilayer film
completely (no data shown). Based on these observations,
the transfer of intact bilayers into the AFM liquid cell
could be verified.
Solubilization of Monomeric 1 in Monolayers and
Bilayers. One of the components of our micropatterning
strategy is to selectively remove monomeric lipids by an
organic solvent or by a detergent solution. In the current
study, we have employed sodium dodecyl sulfate (SDS)
solution as the solubilization agent. Figure 8 shows the
solubilization efficiency of monomeric 1 monolayers and
bilayers from oxidized silicon surfaces as a function of the
concentration of SDS solutions applied. The solubilization
was found to be kinetically controlled and depended on
various factors, such as efficiency of the solution mixing
and immersion time in the SDS solution. However, with
constant conditions (30 min immersion without external
mixing of the solution) we observed a reproducible removal
of 1 from the solid surface. For these controlled conditions,
the removal depended only on the SDS concentration.
Higher SDS concentration removed the film more
effectively under otherwise constant conditions, as expected. However, monolayers and bilayers of 1 showed
markedly different solubilization properties. In the case
of monolayers, significant removal of the film occurred
even at concentrations below the critical micellar concentration (cmc) of SDS (ca. 0.008 M).36 In the case of
bilayers, the film thickness remained unchanged as long
as the SDS concentration was below the cmc. As the SDS
concentration was increased beyond the cmc, the thickness
of the remaining film gradually decreased. Complete
solubilization was achieved if 0.08 M of SDS was used.
Figure 9. Polymerization of 1 by UV irradiation: (a) UV-visible absorption spectra (A) before and (B) after polymerization for
9 min; (b) dependence of UV absorption intensity at 470 nm on UV light irradiation time.
Photopolymerization of Diacetylene Phospholipid
Langmuir, Vol. 19, No. 17, 2003 6999
Figure 10. (a) Morphology of homogeneously polymerized bilayers of 1 after SDS treatment as unveiled by CM-AFM in water.
(b) The lattice of the phospholipid headgroups was resolved in high-resolution CM-AFM images under water. The duration of UV
illumination was 5 min.
Figure 11. Polymerization of 1 by UV irradiation. The bilayer structures were observed by CM-AFM in water after the
photopolymerization and subsequent incubation in 0.08 M SDS solution for 15 h. The UV irradiation time in minutes and the
subsequent treatment are indicated.
Since the polymerized bilayers are stable under these
conditions (vide infra), we have applied this concentration
for the solubilization studies of the partially polymerized
bilayers.
Photopolymerization of Bilayers of 1. Polymerization of 1 by UV light (254 nm) results in the formation
of conjugated ene-yne backbones. This process could be
monitored most conveniently by UV/visible absorption
spectroscopy. Figure 9a shows the UV/visible absorption
spectra of bilayers of 1 before and after polymerization.
The polymeric bilayer has an absorption band between
400 and 550 nm that originates from oligomeric ene-yne
backbones. The intensity of this band as a function of the
UV irradiation time is plotted in Figure 9b. This graph
7000
Langmuir, Vol. 19, No. 17, 2003
Morigaki et al.
Figure 12. (a) CM-AFM height image of patterned bilayer imaged in water; (b) height profile along the white line in (a). The step
height indicated by the markers is 5.9 nm.
Figure 13. Patterned bilayers: fluorescence microscopy images. The scale bars correspond to 50 µm.
allows us to analyze the relation between the UV light
dose and the progress of polymerization. The absorption
intensity increased up to 10 min of irradiation and reached
a plateau; longer irradiation resulted in a slight decline
of the absorption.
The morphological changes in the film during the
polymerization process could be studied by AFM. The
surface of a homogeneously polymerized bilayer of 1, which
may be observed in the patches seen in Figure 10a,
appeared to be as smooth as the unpolymerized bilayers
shown in Figure 5. The lattice of the lipid bilayer could
be resolved in high-resolution CM-AFM images (Figure
10b), which revealed a lattice structure similar to that of
the monomeric bilayers.
The morphological development of the bilayer films as
a function of polymerization time was investigated
systematically. Figure 11 summarizes the observed film
morphologies after various durations of photopolymerization and subsequent treatment with 0.08 M SDS (ca.
15 h). The unpolymerized 1 bilayer was removed completely by the SDS treatment. For the samples with a
short UV exposure, we observed the partial coverage of
the substrate with bilayer domains that were presumably
polymeric. The height difference between the bilayer
domains and the glass substrate was ca. 5.5 nm. As we
(36) Since monolayers have a hydrophobic surface facing the aqueous
solution, their structure is supposed to be energetically less favorable
in water compared with bilayers. However, water alone did not remove
monolayers from the surface, as evidenced by the observation that the
film thickness remained unchanged, if a monolayer was immersed in
water for the same period of time. SDS monomers may adsorb onto the
hydrophobic monolayer surface and cause premicellar aggregation that
solubilizes 1 molecules.
applied a longer UV irradiation time, a larger surface
area was covered by the bilayers (see the image for 9.5
min UV irradiation). However, longer UV irradiation
seemed to induce the following two effects: (i) increase in
the surface roughness; (ii) decrease of the average film
thickness.
Micropatterning. Micropatterning of the bilayers was
achieved by placing a contact lithography mask on the
bilayer surface during the photopolymerization. Figure
12a shows the AFM image of a micropatterned bilayer of
1 that was polymerized for 30 min and treated with a 0.08
M SDS solution. It is evident that the bilayers in the
protected areas (squares), which remained monomeric,
were removed by the SDS solution. The remaining
polymeric bilayer looks heterogeneous with many patches
of slightly higher plateaus. In cross-sectional plots (Figure
12b), we measured a height difference between the
substrate and the remaining film of ca. 5.9 nm.
Fluorescence microscopy under water also showed a
heterogeneous bilayer morphology that resembled the CMAFM data (Figure 13). The square areas did not show
fluorescence at 488 nm, while the surrounding areas
showed fluorescence emission with different intensities.
Upon exposure to air, the polymerized bilayer structure
collapsed partially, as seen in CM-AFM height images
(Figure 14a). However, this morphology is very different
compared to the collapsed monomeric bilayers of 1 (Figure
7). The friction force microscopy image shown in Figure
14b reveals homogeneous friction contrast in the different
areas of the pattern. A similarly well-defined pattern is
observed in tapping mode AFM images (Figure 14c,d).
Photopolymerization of Diacetylene Phospholipid
Langmuir, Vol. 19, No. 17, 2003 7001
Figure 14. AFM images of a patterned bilayer membrane recorded in air. (a) CM-AFM height (z-scale 50 nm) and (b) friction
images (friction force increases from dark to bright). (c) TM-AFM height (z-scale 20 nm) and (d) phase images (phase angle shift
increases from dark to light).
Discussion
In the current study, multimode AFM measurements
have been carried out to characterize the first three steps
of the micropatterning strategy depicted in Figure 1. The
morphology and defect structures were imaged with
tapping mode AFM and contact mode AFM, respectively;
in some cases, the nanometer-scale arrangement of alkyl
tails or lipid headgroups was visualized.
The monomeric Langmuir monolayer of 1 transferred
onto glass substrates had a roughness comparable to the
bare glass substrate and showed few defects. These defects
consisted mainly of dust particles and shallow depressions.
The depth of these depressions (measured as step height
in cross-sectional plots) of ca. 1.3 nm (Figure 3) does not
correspond directly to the monolayer height since it is
unlikely that the bare substrate is exposed at the defects
observed or that the AFM tip can reach to the substrate.
This interpretation is corroborated by the observation of
larger defects in the bilayer structures prepared subsequently by the LS technique. Here step heights of ca.
3.0 nm and ca. 5.6 nm were measured, which can be
attributed to defects in the top leaflet of the bilayer
membrane or to defects of the complete lipid bilayer,
respectively (Figure 5). The same observation was made
when the bilayer structure was intentionally disrupted
in CM-AFM at high force. The experimentally determined
thicknesses agree well with the calculated values based
on the molecular structure (Figure 2), assuming an alltrans conformation of the hydrocarbon tails.
In addition to the thickness measured at defect sites,
the molecular-scale order in the monolayers and bilayers
was found to be consistent with the anticipated structural
organization of the assemblies depicted schematically in
Figure 1. The monolayers exposed the alkane tail to the
air, as can be concluded from the area requirement per
molecule of ca. 29 Å2 deduced from the high-resolution
CM-AFM images (Figure 4a). By contrast, the area
requirement of ca. 66 Å2 agrees well with a tightly packed
array of phospholipid headgroups; thus, in the bilayers
the lipid headgroups are exposed to the aqueous phase
(Figure 6). On the basis of all these experiments, we
conclude that the structure of monolayers and bilayers of
1 deposited onto glass substrates by LB and subsequent
LS techniques is indeed most consistent with the expected
layer structures.
The photopolymerization proceeds without disruption
of the bilayer architecture, as can be concluded from the
CM-AFM data shown in Figure 10, where we find a lattice
7002
Langmuir, Vol. 19, No. 17, 2003
spacing for the photopolymerized phospholipid headgroups
that is similar to the monomers. The AFM observations
of patterned samples (Figures 12 and 14) also have shown
that the polymerized bilayers are resistant to solubilization
by SDS solutions. The dissolution with SDS is kinetically
controlled (vide supra), and the procedure established here
(at concentrations above the solution cmc of SDS) leads
to a complete removal of the unpolymerized, i.e., monomeric, bilayer of 1. The successful removal of the unpolymerized bilayer was independently validated by CMAFM carried out in SDS solution (no data shown) and is
further evident from the AFM data shown in Figures 12
and 14. In the latter images, the collapsed patterned
bilayer has been imaged in air, and a very clear contrast
between the bilayer regions and the empty squares can
be observed. The friction and phase contrast (Figure 14)
are related to the difference in mechanical37-39 and
chemical40,41 properties of the polymerized areas vs the
unpolymerized and subsequently dissolved monomeric
areas (i.e., glass). The high friction forces and the
pronounced phase lag observed in the squares is consistent
with the interpretation that bare glass (high surface
energy) is exposed, and the low friction forces observed in
the surrounding matrix suggest the presence of a low
surface energy layer. The observed features of the polymerized domains are in clear contrast to the case of
monomeric bilayers that show a characteristic collapsed
structure upon exposure to air (Figure 7).
However, the experiments with various UV doses (see,
e.g., Figure 11) suggest that the polymerization either
does not proceed homogeneously or that UV light-induced
degradation of the film competes with the rather slow
polymerization (vide infra). Similar to the AFM data
(Figure 12), where the height differences are directly
measured, bright, i.e., highly fluorescent, patches were
observed by fluorescence microscopy (Figure 13). In the
UV spectra it was evident that longer irradiation times
resulted in a slight decline of the absorption for the band
attributed to the conjugated polymeric backbone. These
(37) Carpick, R. W.; Salmeron, M. Chem. Rev. 1997, 97, 1163-1194.
(38) Schönherr, H.; Vancso, G. J. Mater. Sci. Eng., C 1999, 8-9,
243-249.
(39) Magonov, S. N.; Elings, V.; Whangbo, M.-H. Surf. Sci. 1997,
372, L385-L391.
(40) Frisbie, C. D.; Rozsnyai, L. F.; Noy, A.; Wrighton, M. S.; Lieber,
C. M. Science (Washington, D.C.) 1994, 265, 2071-2074.
(41) Noy, A.; Sanders, C. H.; Vezenov, D. V.; Wong, S. S.; Lieber, C.
M. Langmuir 1998, 14, 1508-1511.
(42) It should be noted that the wavelength and intensity of the UV
adsorption band do not correlate directly to the real degree of
polymerization, since the effective conjugation length is limited by the
local conformation of polymer backbones and the absorption band does
not represent longer polymer chains.
Morigaki et al.
observations might indicate side reactions (degradation
or depolymerization) or a reorganization of the bilayer
films during the photopolymerization process.42
The observation of the solubilization-resistant bilayers
after a very short UV irradiation, as shown in Figure 11,
does not agree with our previous view of the polymerization
process. Previously, we have assumed that the polymerization proceeds homogeneously within the whole bilayer
and the cross-linking occurs gradually after sufficient UV
irradiation.9 By contrast, the new data indicate that parts
of the bilayer become resistant to the solubilization with
a small dose of UV light.
The observed local differences in solubility (Figure 11)
are likely the result of spatially heterogeneous polymerization behavior. Since the illumination intensity can be
assumed to be constant on the relevant sub-λ length scales,
we postulate that the observed local differences are caused
by heterogeneous polymerization due to differences in local
packing; i.e., there is a coexistence of domains with
different molecular organization. Such differences in local
packing or molecular organization would result in spatially
different degrees of polymerization and thus different
resistance to solubilization by SDS.
From the above observations, we can draw the following
summary. To achieve an effective patterning of bilayer
membranes, the basic requirements for the polymerization
process are that the polymerization proceeds to a completion to build a fully cross-linked polymeric network, and
that the bilayer structures remain intact during the
polymerization process. The current AFM observations
suggest that the general bilayer structures of 1 are
preserved during the polymerization process such that
micropatterning of biomimetic membranes based on
lithographic photopolymerization is basically feasible.
However, at a more microscopic level, AFM also revealed
the possible reorganization of the films during the
photopolymerization. It is important for the further
development of micropatterned membranes to understand
how such reorganization affects the properties of polymeric
bilayers as a matrix and as a diffusion barrier for fluid
biomimetic membranes.
Acknowledgment. H.S. gratefully acknowledges financial support by the Deutsche Akademischer Austauschdienst (DAAD) in the framework of the “Hochschulsonderprogramm III” and the NSF MRSEC Center on
Polymer Interfaces and Macromolecular Assemblies (CPIMA) under DMR 9808677.
LA034078F